Exchange of experience and focus on specialisation are key to Europe regaining status in EuroHPC and e-Infrastructures

15 May 2018 Sofia - After the recent e-IRG and Vi-SEEM Workshops, Primeur Magazinewas invited by Ivan Dimov, Deputy Minister of Education and Science at his cabinet in Sofia, Bulgaria for an interview. Ivan Dimov gave a keynote at both workshops about e-Infrastructures and supercomputing. He is responsible for research in Bulgaria. Before that, Ivan Dimov was a scientist, working in computer science. He has been Director of the Institute of Information and Communication Technologies at the Bulgarian Academy of Sciences. He has also been working in some other places in Europe and in America and was for some period of time in England at the University of Reading, in Germany and in France.

Primeur Magazine:And now, you are very busy with the Presidency of the European Union, currently being carried out by Bulgaria. A few weeks ago, there was a conference on EuroHPC which was organized by Bulgaria in collaboration with the European Commission. This conference presented a very nice overview of the state of HPC in Europe and EuroHPC. Can you comment a little bit on that meeting?

Ivan Dimov:This was a very important conference for Bulgaria but also for Europe. In the past years, Europe was not among the ten biggest users of supercomputing, so it was necessary to decide what one should do in order to be again a big player in this area. Without supercomputing, without this tool for researchers, you cannot achieve big results. Even if you deal with theoretical physics, chemistry or biology, you often need such a big machine to perform your computations in order to discover new knowledge. Data mining means dealing with Big Data you are trying to extract knowledge from. This is a big challenge for Bulgaria where there is quite a strong community dealing with high performance computing. Unfortunately, Bulgaria has no big computational power. There is a supercomputer but we need to unite our effort with all the countries in Europe. Commissioner Mariya Gabriel is Bulgarian and responsible for this part of the Commission. It was very important to be able to see what Europe should do to perform better in this important area.

Primeur Magazine:The conference brought this a little bit further. And now, there was a workshop about e-Infrastructures which is also a strong component for high performance computing. How do you see the relationship between high performance computing and e-Infrastructures?

Ivan Dimov:The research infrastructure is extremely important for scientists because without infrastructure it is not possible to deliver high quality results. The normal way is that you work in research projects. You produce innovations and the innovations are going to the industry in order to make the European citizens happy. This is a very important chain of events: from research to innovation to production.

If you compare Europe to the United States of America or China, Europe is not a big player in this area. That is why we need to unite the effort of all the European countries. This is an area where each country has a specific experience and combining this experience for researchers, we reach the best in this area. In Bulgaria, there is quite a strong community dealing with algorithms, software, and middleware. Bulgaria is more specialized in software dealing with algorithms and software which is extremely important for high performance computing. Very often, you have computers with high peak performance but low computational efficiency. One should care about specific algorithms that should be prepared to work on parallel machines. It is not trivial to distribute your tasks or your jobs on a large number of processors and have an efficient implementation. This is a sort of niche in Bulgaria where we focus on and we may contribute to the HPC process. Of course, we are working together with colleagues from other countries. It is a matter of policy of how to organize all this business in order to be more efficient.

It was decided that the European Commission already allocated one billion euro to buy two large supercomputers and also planned to reserve about 9 billion euro for the next Framework Programme in this very important area. At the same time, this infrastructure is also related to electronic infrastructure which is very important for the region of Southeast Europe. This was actually the focus of one of the conferences. If we want to work with countries in the Western Balkans, and this is one of the priorities of our presidency, we need to have good electronic infrastructure. We are putting a lot of effort to fill this gap. This priority is also important for the whole of Europe because you will have smaller discrepancy when you compare research infrastructures in different parts of Europe.

Primeur Magazine:So, basically, it is good to specialize and small countries can also contribute important components but you need to collaborate to achieve results on a European scale.

Ivan Dimov:Absolutely. Modern research areas are very broad. Usually, to compete with big players in this area, you need to specialize in a smaller area to be focused. At the same time, you need to work very closely with colleagues from other countries. Bulgaria has a good experience working with colleagues from Central and Western Europe and also with colleagues from the Balkan Peninsula and Eastern Mediterranean. We need this collaboration and this infrastructure. We can see that people are very active and you have a sort of synergy when you combine different specialists from different countries. This is helpful for the whole research infrastructure in Europe.

Primeur Magazine:Currently, you are working very hard and the focus is on things that are happening during the Bulgarian Presidency. What do you expect in the next year will be important?

Ivan Dimov:The European Commission is about to announce the next Ninth Framework Programme which is a very important task. We are working very hard, also with our colleagues from Austria, who will be in charge of the next Presidency. Europe is not a big player but we know our place and our areas in which we are strong and where we can contribute. We are trying by formulating our national priorities in the field of research to identify a small number of areas where we are strong enough and areas in which we may cooperate. We are focusing on those areas. We have created a number of Centres of Excellence funded by the government in some specific areas where we know that we have good expertise. One of these areas is information and communication technologies. We are very much focused on some specific priorities where we can achieve good results. In such a way, we can say that Bulgarian scientists are good partners in some areas.

There should be a sort of harmony or synergy between national research policy and European research policy. One should identify areas in which it is good to collaborate with big and strong teams. We are also collaborating with teams from America and China. This is absolutely needed. At the same time, we can identify areas that are local areas, connected to heritage and language. We are running projects connected to the digitalization of heritage and texts because we should care about this. At the same time, we have identified areas in which it is good to collaborate on the European level. It is important to have a focused policy to target areas of research where we can expect results.

How we measure this is another important issue. There are metrics which are more or less accepted: using data from databases, in science and other sources to identify areas in which you have papers published in the top one or 10 percent. In such a way, you identify areas in which you can collaborate on a European level and also identify areas that are important for this country. They are local but they also should be stimulated.

Primeur Magazine:In the Netherlands, where I come from, water is an important issue because most of the country is below sea level. Do you have scientific issues which are specific for Bulgaria?

Ivan Dimov:We are focused on genome studies which are connected to food and agriculture because Bulgaria has quite a strong community in this area, also because of some hazards. In the previous year we had a strong snowfall which was a challenge during the month of January. We have very precise models to predict different climate scenarios and weather forecasts, very precise models to be able to predict which part of the highway from Sofia to Burgas should be closed. For this purpose, the supercomputer in Sofia was used, again with very good results. It was convincing politicians that you should put money into this because then you are able to predict with an accuracy of about 300 metres and this was achieved for the first time using such a fine mesh. The mathematical technology is known but the problem was computational power. We had this power at that time. The Bulgarian society understood that ICT is something important and it is good to reserve several millions of euro to buy a good supercomputer to solve this problem instead of becoming stranded in the snow. This was very convincing and after this, we joined the EuroHPC initiative.

Primeur Magazine:Thank you very much for this interview.

Ad Emmen